Immunophenotyping of selected hematologic disorders – focus on lymphoproliferative disorders with more than one malignant cell population

Authors

  • A. Porwit

    Corresponding author
    • Department of Pathobiology and Laboratory Medicine, University of Toronto, University Health Network, Toronto, ON, Canada
    Search for more papers by this author

Correspondence:

Anna Porwit, Department of Pathobiology and Laboratory Medicine, University of Toronto, University Health Network, Toronto, ON, Canada.

Tel.: +1 416 340 3294;

Fax: +1 416 340 5543;

E-mail: anna.porwit@uhn.ca

Summary

Currently, clinical laboratories face increasing demand for flow cytometry testing combined with limited funding. Therefore, many laboratories search for panels that would provide sufficient immunophenotyping information and meet economical requirements. At the Flow Cytometry Laboratory, University Health Network, Toronto, ON, Canada, we apply two 10-color tubes of surface markers for diagnosis of lymphoproliferative disorders (LPDs). These tubes contain most of the mandatory B- and T-cell markers according to European Leukemia Net (www.leukemia-net.org) recommendations. The B-cell-oriented panel includes the following antibodies: Kappa-FITC/lambda-PE/CD19-ECD/CD38-PC5.5/CD20-PC7/CD34-APC/CD23 APC-AF700/CD10 APC-AF750/CD5-PB/CD45-KO. A different combination is applied to detect cytoplasmic Ig light chain expression and aberrant immunophenotype of plasma cells. The T-cell panel allows enumeration of various T- and NK-cell subsets: CD57-FITC/CD11c-PE/CD8-ECD/CD3-PC5.5/CD2-PC7/CD56-APC/CD7-APC-AF700/CD4-APC-AF750/CD5-PB/CD45-KO. The reported overall incidence of B-cell chronic LPDs presenting with more than one aberrant population is approximately 5%. Multicolor analysis facilitates the detection of multiple aberrant populations in the same sample because expression of multiple antigens can be studied simultaneously in each defined population. Examples of LPDs with multiple aberrant populations are presented.

Introduction

The recently published EuroFlow antibody panels for immunophenotyping of B-cell and T-cell lymphoproliferative disorders (LPD) recommend using 42 antibodies in the B-cell panel (five tubes, one with 11 antibodies, three with eight antibodies, and one with seven antibodies) and 43 antibodies in the T-cell panel (four tubes with eight antibodies, one with seven, and one with five) [1]. These panels were developed to reproduce as much as possible the WHO 2008 classification diagnostic ‘gold standard’ solely based on flow cytometry (FCM) [1, 2]. However, in clinical practice, the diagnosis is seldom made solely based on FCM. Flow cytometry is used as a part of integrated diagnostics combining morphology, immunophenotyping, and genetic findings [2, 3]. Currently, clinical laboratories face increasing demand for FCM testing combined with static/declining funding or limited reimbursement. Therefore, many laboratories search for panels that would provide sufficient immunophenotyping information and meet economical requirements.

European Leukemia Net Work Package 10 (ELN WP10, ‘Diagnostics’, www.leukemia-net.org) has published recommended panels based on the consensus collective view of participants of the ELN WP10 group as to the numbers and characteristics of markers necessary for diagnosis of hematologic malignancies by FCM [4]. For diagnosis of lymphoproliferative disorders, a panel of antibodies is proposed [4] divided into B-cell, T-cell, and NK-cell-oriented subsets of antibodies (markers in bold are considered here in as B-cell, T-cell, and NK-cell related, respectively):

  • B-cell-oriented panel: CD19, k, l, surface Immunoglobulin (Ig) heavy chains, CD5, CD10, CD20, CD22, CD23, CD25, CD38, CD103, CD79b.
  • T-cell-oriented panel: CD3, CD2, CD4, CD5, CD7, CD8.
  • NK-cell-oriented panel: CD56, CD16, CD45, CD57, CD94, CD158.

Markers listed as additional include FMC7, CD11c, CD38, CD43, CD81, and CD123 [4].

ELN WP10 document suggests that some degree of harmonization in the panel choice should be reached even considering the increasing complexity of multiparameter FCM. However, it is advisable that some freedom and flexibility remain in the choice of methodology in various laboratories, because similar results may be obtained using various combinations of markers and various flow cytometers.

At the Flow Cytometry Laboratory, Department of Laboratory Medicine, University Health Network, Toronto, ON, Canada, we apply two main 10-color tubes of surface markers (one B-cell and one T-cell oriented, see below) for diagnosis of lymphoproliferative disorders. The two tubes contain most of the mandatory B- and T-cell markers according to ELN recommendations. Due to low incidence of pure NK-cell lymphoproliferations, we did not include all recommended NK-associated markers.

Additional combinations are applied for hairy cell leukemia diagnosis (one 6-color tube) and to further characterize samples with suspected T-cell lymphomas (one 10-color tube). A separate one 10-color tube with a combination of surface and cytoplasmic markers has been created for samples from patients with monoclonal protein in serum electrophoresis for diagnosis of plasma cell dyscrasia and lymphoplasmacytic lymphoma. Acquisition is performed on Navios flow cytometer, and Kaluza software is used for analysis (Beckman Coulter, Miami, FL, USA). The 10-color FCM program has been validated against the previously applied 5-color analysis and accredited according to the Ontario Laboratory Accreditation requirements.

Analysis of B- and T-cell Subsets with Multicolor Approach

B-cell malignancies are clonal expansions of B-cells that express only one type of Ig light chain (κ or λ) or (rarely) lack surface Ig light chain expression. However, monotypic expression of Ig light chain is not equal to neoplasia because it has been reported in some reactive conditions {reviewed in ref. [3]}. Also, abundant, reactive, polyclonal B-cell populations may make detection of smaller neoplastic monotypic B-cell populations difficult. Multiparameter 8–10 color FCM allows analysis of light chain expression in total CD19 +  B-cell population and also in various subsets of B cells such as CD5+/CD19+, CD23+/CD5+/CD19+ or CD10+/CD19+ positive cells. In analysis of bone marrow samples for lymphoma staging, it is also important to evaluate normal B-cell differentiation [5], which allows detection of subpopulation of cells with aberrant expression of CD20, CD10, CD34, and CD38.

The B-cell-oriented antibody combination applied at UHN includes the following antibodies:

  • Kappa-FITC/lambda-PE/CD19-ECD/CD38-PC5.5/CD20-PC7/CD34-APC/CD23 APC-AF700/CD10 APC-AF750/CD5-PB/CD45-KO1

Analysis of B-cell subsets in a bone marrow sample is exemplified in Figure 1a. Samples are analyzed first using CD45-Side scatter (SSC) plot to map main hematopoietic cell populations. Further analysis is focused on CD19+ B-cell population gated on CD19-SSC plot. B-cell differentiation pattern is evaluated on CD34/CD38 and CD20/CD10 plots. These plots are also used to detect mature B-cell populations with aberrant CD20 and/or CD38 expression. Early precursor, intermediate precursor, and mature B-cell compartments are enumerated. Kappa and lambda light chain expression is evaluated separately on total CD19+ population and on various subsets of B-cells (Figure 1a). Separate plot gated on CD20 is created to detect B-cell LPDs with aberrantly low CD19 expression (Figure 1a). If a CD11c-positive B-cell population is detected in T-cell-oriented tube, a hairy cell leukemia panel (CD103 FITC/CD11c-PE/CD19-ECD/CD123 APC/CD25 PB/CD45-KO) is added.

Figure 1.

Ten-color analysis of lymphoproliferative disorders. (a) Staging bone marrow from a patient with follicular lymphoma was analyzed with the B-cell panel Kappa-FITC/Lambda-PE/CD19-ECD/CD38-PC5.5/CD20-PC7/CD34-APC/CD23 APC-AF700/CD10 APC-AF750/CD5-PB/CD45-KO (all antibodies Beckman Coulter, Miami, FL, USA). Analysis shows good viability and 10.45% cells in the lymphocyte gate (left upper plots). B cells were 7.9% of bone marrow cells, and a population of CD20 dim/CD10+ B-cell precursors was detected (right upper plots). Analysis of kappa and lambda expression in the total B-cell population shows kappa/lambda ratio 0.5 in the SIg-expressing B cells. However, when a population with aberrantly low CD19 expression, together with strong CD20 expression and weak CD10 expression, is analyzed (right lower plots), the kappa/lambda ratio is 0.1. The aberrant population was CD38 dim, while B-cell precursors displayed normal strong CD38 expression and were in part CD34 positive (not shown). (b) Bone marrow sample from a patient with paraproteinemia was analyzed with cyt.kappa-FITC7/cyt. Lambda-PE/CD56-ECD/CD138-PC5.5/CD117-APC/CD19 – APC-AF700/CD38-APC-AF750/CD20 – PB/CD45-KO panel (kappa and lambda antibodies from Dako, Glostrup, Denmark, other antibodies from Beckman Coulter, Miami, FL, USA). 10% of the cells were plasma cells (CD38/CD138 bright) with monotypic expression of lambda (right upper plots, blue dots). Plasma cell were CD45 dim, CD117 dim, CD19 dim, 20 positive, CD56 negative, and had intermediate scatter characteristics (left upper and middle plots, blue dots). CD19+ lymphocytes were 18% of bone marrow cells and showed monotypic expression of kappa (middle right plots, green dots). The kappa-positive B-cell population had B-CLL-related immunophenotype (CD23+, CD5 dim) but showed higher CD20 positivity than usually seen in B-CLL (lower plots, green dots). (c) Blood sample from a patient with lymphocytosis was analyzed with T-cell-oriented panel: CD57-FITC/CD11c-PE/CD8-ECD/CD3-PC5.5/CD2-PC7/CD56-APC/CD7-APC-AF700/CD4-APC-AF750/CD5-PB/CD45-KO. 53% of cells were in the lymphocyte region (upper left plot). Most of the lymphocytes were CD3-, CD2-, and CD5-positive T-cells (48% of blood cells). A population of CD3+, CD57+, CD2+, CD5+, CD4+, CD56-, CD7-, CD8-, CD11c (red dots) lymphocytes was detected (50% of lymphocytes, 25% of blood cell). T-cell receptor rearrangement study confirmed the presence of a clonal T-cell population.

In patients with known paraproteinemia, a different tube is applied instead of the B-cell tube to detect cytoplasmic (cyt.) Ig light chain expression and aberrant immunophenotype of plasma cells:

  • cyt.kappa-FITC7/cyt. Lambda-PE/CD56-ECD/CD138-PC5.5/CD117-APC/CD19 – APC-AF700/CD38-APC-AF750/CD20 – PB/CD45-KO.

In this panel, the cytoplasmic Ig expression is evaluated separately in the CD19+/CD20+ population with lymphoid scatter and strong CD45 expression (B lymphocytes) and in CD38++/CD138+ plasma cell population (Figure 1b). Expression of CD56, CD117, CD45, CD19, and CD20 is then evaluated in plasma cells. If a monoclonal B-cell population is detected, the surface B-cell tube is also run for further immunophenotyping of B-cells. Figure 1b shows a bone marrow sample with coexistent monoclonal plasma cell population with aberrant CD117 expression and monoclonal B-cell population with B-CLL-associated immunophenotype.

A T-cell-oriented tube applied at UHN is aimed on detection of aberrant marker expression and enumeration of various T-cell and NK-cell subsets:

  • CD57-FITC/CD11c-PE/CD8-ECD/CD3-PC5.5/CD2-PC7/CD56-APC/CD7-APC-AF700/CD4-APC-AF750/CD5-PB/CD45-KO.

Analysis of T-cell tube is exemplified in Figure 1c. T- and NK-cell subsets are analyzed using gating on lymphocytes (CD45 bright), on CD3, CD2+/CD7+, CD56+, and CD57+ cell populations. Multicolor approach allows enumeration of various subsets of T and NK cells that may increase in infectious and inflammatory conditions, as well as detection of aberrant populations with decreased expression of some T-cell-associated antigens. Subpopulations of T or NK cells that are present in reactive conditions, if clonally expanded, may represent a malignant cell population. Therefore, the results have to be correlated with clinical data, cytogenetic analysis, and molecular analysis of T-cell receptor rearrangement. For example, CD3+/CD57+/CD8+ cells often increase in viral infections but also may represent a manifestation of large granular lymphocyte leukemia [6]. CD57+ CD4+ lymphoproliferations are even more rare (Figure 1c) [7]. CD3+ CD4+ CD8+ cells increase in autoimmune conditions and in viral infections but also may represent a leukemic cell population in T-cell prolymphocytic leukemia [8]. Evaluation of several antigen expression on given cell subsets allows more precise differentiation between normal and aberrant populations. In samples where a population of cells suspect for manifestation of T-cell lymphoma is found and in lymph node cell suspensions, an additional tube is also analyzed for further immunophenotyping of the aberrant T-cell population and classification of T-cell lymphoma according to the WHO classification:

  • CD30-FITC/TCR alpha/beta-PE/CD8-ECD/TCD gamma/delta-PC5.5/CD3-PC7/CD1a-APC/CD7 APC-AF700/CD4 APC-AF 750/CD25-PB/CD45-KO.

However, in the majority of T- and NK-cell lymphomas presenting in lymph nodes, FCM analysis is not sufficient, and further immunophenotyping by immunohistochemistry allowing correlation of cell morphology with specific immunophenotype is necessary for the final classification.

Lymphoproliferative Disorders with More Than One Aberrant Cell Population

The overall incidence of cytogenetically unrelated clones in hematologic malignancies in a series of 1110 cases was 2.4% [9]. Patients presenting simultaneously with composite B- and T-cell lymphomas are very rare. In this group, angioimmunoblastic T-cell lymphomas with diffuse large B-cell lymphomas showing EBV positivity in the B-cell component are the most common [10, 11]. Occurrence of T-cell lymphoma in patients with B-cell chronic lymphocytic leukemia (CLL) has also been reported [12-14]. Conversely, patients with T/NK Large granular lymphocytic LPDs often present also with monoclonal B-cell lymphocytosis (MBL) [15].

Several case reports of biclonal or composite B-cell lymphomas have been published [16-22]. The reported overall incidence of B-cell chronic LPDs presenting with more than one aberrant population in one larger series (n = 477) studied by FCM was approximately 5% [23]. Multicolor analysis facilitates detection of multiple aberrant populations in the same sample because expression of multiple antigens can be studied simultaneously in each defined population. The most common B-cell LPDs with more than one aberrant cell population are those where both populations display B-CLL phenotype, but one is kappa positive and the other lambda positive (34% of biclonal cases in the above-mentioned series) [20, 23-25]. Often, one of the populations has immunophenotypical features associated with the so-called atypical CLL such as strong CD20 expression [23]. An example of such case is illustrated in Figure 2a showing two CD19+/CD5+/CD23+ populations, one positive for kappa, with stronger CD20 expression and positive for CD38 and one positive for lambda, with weaker CD20 expression and negative for CD38. Rare cases of marginal or lymphoplasmacytic LPDs with one kappa-positive and one lambda-positive population have also been described [18, 26]. Molecular studies in most LPDs with one kappa- and one lambda-positive population with B-CLL-associated immunophenotype revealed two separate clonal B-cell populations with different patterns of IgH rearrangements. However, in one reported case of t[14, 18] (q32;q21), positive follicular lymphoma presenting with a kappa-positive population in the bone marrow and a lambda-positive population in the lymph node, the tumor clones shared an identical BCL2-IgH recombination and identical variable, diversity and joining segments together with clone-specific VH somatic hypermutations on the untranslocated IgH allele. This study suggests that kappa- and lambda-positive clones in this patient developed from a common progenitor [21].

Figure 2.

Examples of samples with two aberrant B-cell populations: (a) Analysis of blood from a patient with lymphocytosis using B-cell tube showed 44% of cells in the lymphocyte region and 22% B cells. Kappa/lambda ratio in the total B-cell population was 0.4. Further analysis revealed that both kappa- and lambda-positive cells were positive for CD5 and CD23. Kappa-positive population had stronger CD20 expression, weaker CD5 expression and was positive for CD38 (green dots), while lambda-positive population had weaker CD20, stronger CD5 and was negative for CD38 (red dots). (b) Analysis of blood from a patient with peripheral lymphadenopathy showed 34% of lymphocytes and 11% of B-cells. Two populations of B cells were noted: one with stronger CD20/CD19 expression and larger according to the scatter characteristics (blue dots) and the second with weaker CD20/CD19 expression and smaller (pink dots). Most cells with stronger B-cell markers expression were negative for CD5 and CD23 and positive for lambda. This population was also positive for CD11c (not shown). Cells with weaker expression of B-cell markers were positive for CD23, had weaker partial expression of CD5, and no expression of kappa or lambda light chains. Subsequent biopsy of a lymph node showed composite lymphoma with areas of small lymphocytic lymphoma and areas of marginal zone lymphoma.

Another group of cases with more than one aberrant population is characterized by the presence of two morphologically and immunophenotypically different B-cell LPDs. Most often, one of the populations has a B-CLL immunophenotype [23]. Also, most often, one of the populations is present at low frequency, suggesting B-CLL MBL associated with another B-cell LPD or non-CLL MBL associated with B-CLL [23]. The other population may have immunophenotypic characteristics of lymphoplasmacytic lymphoma, marginal zone lymphoma, hairy cell leukemia, follicular lymphoma or mantle cell lymphoma. Figure 2b illustrates a case where two aberrant CD19+ B-cell populations were found at low frequency in peripheral blood: one population with B-CLL immunophenotype (CD5+, CD23+, CD11c−, negative for SIg) and the other with marginal zone lymphoma–associated immunophenotype (CD5−, CD23−, CD11c+, Lambda+). Subsequent biopsy of enlarged lymph node confirmed the presence of both a marginal zone lymphoma and a small lymphocytic lymphoma.

Rare cases of composite mantle cell lymphoma and follicular lymphoma have also been reported [19]. Molecular analysis revealed two separate clonal populations in most studied cases of immunophenotypically distinct LPDs [17, 19]. LPDs, mostly of B-CLL immunophenotype have also been infrequently reported in patients with plasma cell myeloma (Figure 1b). In most cases studied by molecular methods, two separate clonal populations were found [27]. However, in rare patients, a common clonal origin was suggested [22, 28].

Of note, biclonal populations have also been found in patients with MBL both of B-CLL and non-CLL immunophenotype [29, 30]. The frequency of biclonal populations in MBL seems to be much higher than in patients with LPDs (approximately 20%), suggesting oligoclonal diversification associated with antigen-driven expansions [30].

Note

  1. 1

    FITC - Fluorescein isothiocyanate, PE- R-Phycoerythrin, ECD – PE-Texas red-X, PC5.5 – PE-Cyanine 5.5, PC7 - PE- Cyanine 7, APC- Allophycocyanin, APC-AF700 – APC- Alexa Fluor 700, APC-AF750 - APC-Alexa Fluor 750, PB – Pacific Blue, KO – Krome orange.

Ancillary